When AI scrapes your art
Summary
This episode examines the legal and economic impact of generative AI on artists, featuring Kelly McKernan, whose work was scraped to train image models. Host Dina Temple Raston discusses how AI imitation threatens livelihoods and explores countermeasures like "poisoning" data. Legal expert Jason Schultz analyzes the "fair use" doctrine and the Anderson v. Stability AI case, noting that US courts currently favor AI training over copyright claims.
Topics discussed
Intro: Can generative AI be creative or just copy?
Sponsor segments: Nord Security and Serval
Context: Artists vs. AI and the legal battle
Meet Kelly McKernan: A 15-year career in art
Discovery: Kelly's work scraped for AI training
The threat: AI mimicking style and losing clients
Impact: Income drop and emotional reaction
Realization: Friends spotting the AI-generated art
The problem: Lack of consent and permission
Resistance: Artists 'poisoning' work to confuse AI
Legal gap: Copyright vs. style and consent
Case status: Kelly's claims dismissed, seeking remedy
Listener mail: Speed of AI vs. human creativity
Broader context: Tech calls to slow AI development
Kelly's next step: Returning to school for MFA
Sponsor segments: Odoo and Hunting Warhead
Legal analysis: Fair use and infringement in AI
Consumer perspective: Can we legally ban AI art?
Regulation: Proposed laws and attribution issues
Fair use details: Public availability and scraping
International law: US vs. Europe AI regulations
Anderson v. Stability AI: Potential verdict impact
Parallels: Meta settlement vs. Big Tobacco deal
Credits, archive info, and Recorded Future sponsor
Listen ad-free on Castria